IMPLEMENTATION NOTE

Integration of Recurrent and Capital

“Development” Budgets:
Issues, Problems, Country
Experiences, and
the Way Forward

Feridoun Sarraf
World Bank Consultant

July 2005

The World Bank

Public Expenditure Working Group

Prepared with funding support from the Public Expenditure

and Financial Accountability (PEFA) Program
Abstract
Dual budgeting systems persist in developing countries. This paper identifies and
describes both domestic institutions and donor practices that have been contributing to
separate recurrent and development budget management systems in different stages of the
budget cycle. It examines both the causes of resistance to change and pays attention to the
institutional dynamics that may make integration of the recurrent and development
budgets possible. Drawing from the experience of the few developing countries that have
made progress in this area, it offers practical guidance for countries and other
stakeholders on a full integration of recurrent and development budgets, including the
institutions and the management of budget preparation, execution, and reporting of a
unified budget system. Recommendations are made for country authorities and a possible
donor-coordinated policy dialogue with client countries and associated technical support.

This paper is part of a larger effort by The World Bank and its PEFA partners to focus more
on public financial management capacity development, including the development of
guidance and mainstreaming of lessons on specific issues of concern in implementing
reforms. Public financial management includes all phases of the budget cycle, including the
preparation of the budget, internal control and audit, procurement, monitoring and reporting
arrangements, and external audit. The broad objectives of public financial management are
to achieve overall fiscal discipline, allocation of resources to priority needs, and efficient
and effective allocation of public services.

The Public Expenditure Working Group is a joint effort of The World Bank’s Financial
Management, PREM, and Procurement Networks, with the IMF and PEFA Secretariat, to
support improvement in country public financial management system performance. For
more information, see
www1.worldbank.org/publicsector/pe/StrengthenedApproach/index.htm

PEFA is a partnership between the World Bank, the European Commission, the UK's
Department for International Development, the Swiss State Secretariat for Economic
Affairs, the French Ministry of Foreign Affairs, the Royal Norwegian Ministry of Foreign
Affairs, the International Monetary Fund and the Strategic Partnership with Africa. For
more information, see www.pefa.org

ii
Abbreviations and Acronyms
AAP Assessment and Action Plan
CFAA Country Financial Accountability Assessment
DFID Department for International Development
GFSM Government Financial Statistics Manual
HIPC Highly Indebted Poor Countries
IFMIS Integrated Financial Management Information System
IMF International Monetary Fund
MTEF Medium-Term Expenditure Framework
OECD Organization for Economic Co-operation and Development
PER Public Expenditure Review
PIP Public Investment Program
PRSP Poverty Reduction Strategy Paper
ROSC Report on Observance of Standards and Codes
SNA System of National Accounts

iii
Contents

Abstract ............................................................................................................................... ii
Abbreviations and Acronyms ............................................................................................ iii
1. Introduction and context ............................................................................................. 1
2. The roots and institutional development of dual budgeting........................................ 3
Origins............................................................................................................................. 3
Dual budgeting in developing countries ......................................................................... 5
1960s to late 1990s: post-independence, construction, national development
planning, and public investment program era............................................................. 5
Late 1990s to date: PRSP, MTEF, sector-wide approach, and budget support era .... 7
3. Budget integration: four necessary elements .............................................................. 8
Organizational and staffing integration into a single ministry........................................ 9
The way forward ....................................................................................................... 11
Integration of budget preparation.................................................................................. 11
The way forward ....................................................................................................... 13
Integration of budget documentation and presentation................................................. 14
The way forward ....................................................................................................... 15
Unified budget execution, accounting, and reporting systems ..................................... 15
The way forward ....................................................................................................... 17

iv
1. Introduction and context 1

A familiar symptom of weak public financial management in developing countries,

identified in the public finance development literature, is completed capital projects that
lack the recurrent funds they need to function properly: hospitals that lack funding for
medical staff and supplies; schools with no funding for teaching staff, supplies, and
books; or roads with no funds in the government’s budget to maintain them. Though the
calculation and budgeting for recurrent costs of physical assets created by the government
budget is a well established budgeting principle, the problems arise from several causes:
poor planning and poor information sharing across different processes; lack of planning
for future spending within resource constraints, and inability to coordinate donor-
financed and domestically-financed budgetary operations. In recent years, the extension
of developing countries’ capital budgets to include, in addition to capital outlays,
substantial amounts of “developmental” recurrent expenditure mostly financed by donor
funds, has complicated matters further, and concerns are frequently raised about the weak
interaction of the widened capital budget or “development budget” with a government’s
traditional recurrent budget.

Existing dual budgeting, accounting, and reporting systems do not provide for a
comprehensive analysis in the planning and budgeting of government operations or in the
evaluation of their outcomes and results. Double budgeting for the same recurrent activity
from separate sources may well happen. Moreover, government accounting and banking
rules may differ not only between recurrent and capital budgets, but also between two
recurrent budgets, one financed from domestic and the other from external sources.

Recent initiatives and developments in public expenditure management call for a

renewed emphasis on the integration of recurrent and development budgets. These
initiatives include reliance on medium-term expenditure frameworks (MTEF), the
preparation of poverty reduction strategy papers (PRSPs), the identification and
monitoring of poverty reducing expenditures in developing countries’ budgets and
financial reporting systems, and the introduction of sector-wide and general budget
support approaches along with traditional support for individual projects by most donors.
Experience shows that when the preparation and execution of recurrent and development
budgets are institutionally, organizationally, and technically separate, real coordination
between the two is nearly impossible.

The separation of recurrent and development budgets can occur along several
dimensions: (i) separate central ministries managing processes and making different
decisions , (ii) separate planning and budget allocation processes, (iii) separate budget

1
The author wishes to thank members of the Public Expenditure Working Group, particularly Bill
Dorotinsky, Richard Allen, Tej Prakash, Mike Stevens, Nicola Smithers, Odile Keller, Bernard Myers and
Frans Ronsholt, and Julie Lynn of the DFID for their helpful comments for preparing this paper. Editorial
assistance was provided by Rachel Weaving. Remaining errors are the author’s.

1
documentation processes; and (iv) separate budget execution, accounting, banking, and
reporting processes.. Several different terms are in use, such as “recurring” and
“nonrecurring”, “ordinary and extraordinary”, “revenue and capital”, “current and
capital” , “current and investment”, “above and below the line”, “operational and
development” and “recurrent and development” budgets. This paper uses the terms
“recurrent and development budgets”, because most developing countries currently use
these terms. Today in developing countries, dual budgeting is manifested by domestically
financed recurrent expenditures, called the “recurrent or current budget”, as distinct from
what is jointly called the “capital or development budget”, consisting of externally-
financed capital—and, increasingly, recurrent—expenditures, plus expenditures financed
from government counterpart funds (recurrent or capital), plus the government’s own-
financed capital expenditures. 2

Almost all international and several bilateral providers of technical and financial
assistance have been advising their client countries to take steps toward integrating their
recurrent and development budgets. For example, the World Bank’s Public Expenditure
Management Handbook describes the separation of recurrent and development budgets as
the most important culprit in the failure to link planning, policy, and budgeting, and in
poor budgetary outcomes. 3 The Bank’s “Country Financial Accountability Assessment
(CFAA) Guidelines to Staff” require asking government officials whether any separate
investment budget is adequately integrated with the recurrent budget.4 And the
International Monetary Fund (IMF) warns its fiscal economists that a dual budgeting
system may be based on inconsistent macroeconomic assumptions, budget classifications,
and accounting rules. 5 Other World Bank/IMF advisory notes and/or working documents
raise the same issues and concerns. 6 Finally, most country-specific public expenditure
reviews by the World Bank, and individual technical assistance reports by the IMF,
strongly advise the integration of the recurrent and development budgets in client
countries.

Developing countries’ progress in integrating their recurrent and development budgets

has been very limited. One factor that helps maintain the status quo is antiquated

2
Unlike “capital expenditure”, which has a clear and universal definition in the national accounting system
and economic classification of government expenditures, the term “development expenditure” is only a
conventional and not even universally accepted term, although it has been widely in use since the 1960s in
developing countries.
3
Public Expenditure Management Handbook, World Bank, 1998, p. 53.
4
CFAA Guidelines to Staff, World Bank, 2003, p. 18.
5
Guidelines for Public Expenditure Management, International Monetary Fund, 1999, p.19.
6
See, for example, Fiscal ROSC Code and Manual of the IMF (2001); Country Assessment and Action
Plan for HIPCs (HIPC-AAP) Guidelines, jointly prepared by the World Bank and the IMF (2003); the
OECD-SIGMA’s Managing Public Expenditures--A Reference Book for Transition Countries (2001); and
the UK Government Department for International Development (DFID) Guidelines on Understanding and
Reforming Public Expenditure Management (2001).

2
economic development theory, emphasizing capital spending for faster economic growth.
Another is institutional incentives. Donor practices have also tended to reinforce dual
budgeting practices. Donors have traditionally focused on capital investments, and a
desire to attract donor funding gives a country a strong incentive to maintain a separate
development budget process, even though donors’ use of budget support over project aid
and multi-donor involvement in financing recurrent expenditures may increasingly
change that incentive. The failure to ensure resources for ongoing maintenance of capital
investments gives some agencies an incentive to seek earmarked or extra-budgetary
revenues (e.g. road funds). On the other hand, some ministries pursue activities through
the development budget that would otherwise be treated as recurrent spending, due to the
budgetary incentives that are at work. Finally, the simple dynamics of separate ministries
for finance and for development or planning, with their own domestic constituencies, and
governments’ general reluctance to reduce the number of ministerial portfolios, also work
against the integration of recurrent and development budgets.

It seems that launching a new and donor-coordinated approach is needed to address these
barriers. Such an approach is proposed in this paper. 7 Chapter 2 examines the origins and
evolution of dual budgeting systems, paying attention to institutional dynamics that may
hinder or make integration possible. Chapter 3 looks at current problems and the causes
of resistance to change. Drawing on country experience, it offers practical
recommendations for achieving budget integration and for a possible donor-coordinated
policy dialogue with client countries, with associated technical support. 8

2. The roots and institutional development of dual

budgeting
Origins

Dual budgeting originated in European countries, but in those countries it lasted only for
a short period. It was introduced in the late 1930s in order help governments ensure that
the resources they borrowed were used only for capital expenditures. 9 After the Second
7
This paper takes its cue from an earlier paper prepared for the Public Expenditure Working Group. On the
issue of integration of recurrent and capital budgets, that paper suggested that: “the Bank and bilateral
donors could usefully give a lead once again by taking the incremental recurrent costs of the projects they
finance seriously. They should continually emphasize in the dialogue with governments the need to fund
existing policies adequately before investing in new capital assets. Fortunately the progressive shift to
budget support should help integrate better the two budgets.” See Mike Stevens, Institutional and Incentive
Issues in Public Financial Management Reform in Poor Countries, October 2004, pp. 11-13.
8
The integration of recurrent and development budgets cannot solve other possible public expenditure
management problems such as weak foreign aid tracking and reporting or the presence of large off-budget
transactions in a country’s budget system. It remains, however, the only viable option for ensuring the
coordination of these two budgets. As always, the role of supporting reforms should not be underestimated.
9
Detailed information on the creation of dual budgeting in the UK and Sweden in 1930s and 1940s can be
found in A. Premchand, “Government Budgeting and Expenditure Controls--Theory and Practice”,
International Monetary Fund, 1983, pp. 292-302.

3
World War, as governments relaxed their use of borrowed funds, budgets were
integrated. The change in approach reflected several factors: massive postwar
reconstruction work and increased recurrent expenditures, along with the acceptance of
the Keynesian model of linking government spending and the size of the budget deficit
and borrowing requirements to both fiscal and monetary determinants and the business
cycle. 10 Moreover, it soon became clear that the need to reap a return—whether financial,
social, or economic—applied to the entire spectrum of government spending. 11 Hence it
came to be accepted that regardless of their financing sources, government’s recurrent
spending and capital investment are complements in any logical combination that may be
required, and that the two types of expenditures together produce results, provided the
context is one of overall macro-fiscal balance. 12

The United Nations’ System of National Accounts (SNA) and the IMF’s Government
Finance Statistics Manual (GFSM) prescribe how a government’s consumption and
investment activities should be separately recorded in the government budget and in
accounting records and statistics. But they do not call for a dual budgeting system.
Separate data on government consumption and investment can be maintained within a
unified government budget.

Today both advanced and developing countries use government borrowing, along with
government revenues, to finance both recurrent and investment expenditure. What
matters is the overall size of the budget deficit, in the context of macroeconomic stability
and sustainability in a non-inflationary environment. Borrowing has no relation with the
allocation of the borrowed funds to recurrent or capital expenditures, except where a loan
agreement is associated with a specific capital project. 13

10
On the liberating ideas of John Maynard Keynes, and his advocacy of deficit spending, among numerous
resources, see A. Premchand: "Government Budgeting and Expenditure Controls--Theory and Practice”,
International Monetary Fund, 1983, PP. 4-10
11
The UK Government, in an effort to keep its budget deficit in a fiscally sound range, has in recent years
reintroduced the golden rule of limiting capital expenditures to the size of borrowing, but this should be
regarded as a self-imposed fiscal discipline measure only, and has not created any type of dual budgeting.
12
“In the United States, periodic recommendations were made for the introduction of a separate capital
budget. But this never materialized, primarily because it might tilt the resource allocation in favor of
‘bricks and mortar’. However, the budget documents presented a special analysis of investment
expenditures, which was for information only and had no accounting or other implementation impact for
the budget structure.” For further details see A. Premchand, op. cit., p.302.
13
Unlike in developing countries, budgeting terminology in advanced countries uses the term “capital
budget”, rather than “development budget” as used here. In most OECD countries, capital projects have
already been completed over the last few decades and today their maintenance costs are of a current
expenditure nature, along with other recurrent expenditures, including mainly social payments. This makes
the share of capital spending a very small part of advanced countries’ overall spending. The dwindling of
capital spending as a share of total spending is a feature increasingly seen in developing countries, but in
the latter group, the reason is increased necessary recurrent expenditures only; these countries’ need for
capital spending by government will remain high for a long time.

4
Dual budgeting in developing countries

In developing countries, the content and coverage of recurrent and capital expenditures
differ from both the pre-World War 2 meaning (using borrowing only for capital
expenditures) and from SNA and GFS concepts. Though the size of a government’s
budget deficit and borrowing has no necessary relation to the size of its capital
expenditures, dual budget preparation, documentation, and presentation, as well as dual
accounting and reporting systems, remain common features. What are the historical and
institutional reasons for this?

1960s to late 1990s: post-independence, construction, national development

planning, and public investment program era

During the 1960s, developing countries’ nascent institutions of political and economic
management tended to adopt the technical practices of the western democracies. For
example, most Anglophone African countries used dual budgets, with an “above-the-line
account” showing revenues and mainly “recurrent” expenditures, and a “below-the-line
account” showing “extra-budgetary”, interest and lending to nationalized industries and a
very few local governments, where the latter implemented a central government’s
investment plans. 14 Many countries in Africa, Latin America, and the Middle East
adopted national development planning, which had European roots and was still being
practiced in Belgium, France, and Spain. Inexperienced finance ministries could not carry
out the new tasks of medium-term development planning and capital project appraisal, so
planning or development or economic affairs ministries were set up for the purpose. In
practice, the new ministries soon became responsible for the identification, appraisal,
budgeting, and even accounting and reporting of investment budgets, in an initial, but
major and lasting, step toward dual budgeting.

Frequently, finance and development ministries issued their own separate budget
circulars, and the dual approach to budgeting also took root in the line ministries.
Politically visible gaps in physical infrastructure, and then capital project supremacy in
government budgeting, helped to create new financial power bases, leaving the finance
ministries responsible for budgeting civil service salaries and the needs of some small
government ministries with minimal operational expenditures. Lack of coordination
between the finance and planning or development ministries may also have reflected
differences between their ministers and/or heads of state who, in some cases, exercised de
facto control of investment decisions and their funding. Such settings made it difficult to
introduce consolidated budget presentation and classification systems. Moreover, some

14
The “below-the-line account” used by the UK before the 1960s differs from the “below-the-line” concept
used today in some fiscal terminologies (mainly verbal rather than written) for defining the balance of
government operations and how its deficits are financed. Additionally, the term “below-the-line accounts or
funds” also is used in some Anglophone African countries for another meaning. i.e. transferring some
amounts from a ministry’s budget to a special fund, at times supplemented by earmarked revenues and
spending them outside government mainstream accounting and reporting procedures, including different
classifications, carryover of funds from one year to another, minimum or no reporting, etc.

5
substantial but unforeseen recurrent costs evolved from the expansion of investment
projects. 15

International lending institutions and the donor community at large found the new
ministries more capable than weak finance ministries of conducting business in the area
of capital spending, as well as so-called developmental recurrent spending. The donors’
responses further institutionalized the practice of separate organizations preparing
separate budgets. Though the public investment program (PIP) era in the 1980s saw the
introduction of more coordinated project selection and appraisal processes, replacing
individual project appraisal, it did not change the dynamics of dual budgeting. Moreover,
additional donor support to recurrent developmental and humanitarian expenditures
created further coordination problems with governments’ traditional recurrent budgets.

Both before and during the PIP era, emphasis was placed on calculating the recurrent
spending that capital projects would need after their completion, but in practice finance
ministers were uninformed and reluctant or financially unable to provide the necessary
funds. The lack of provision for recurrent funds added to the concerns that had already
been created by the organizational separation of the two budgets. It might be argued that
the main reason for under-budgeting of recurrent expenditures for completed capital
projects was, and still is, general fiscal stress in developing countries. But a counter-
argument is that the fiscal stress could have been projected before the projects were
initiated. Alternatively, it might be said that other spending priorities always crowded out
the claims of these projects on recurrent budgets. Both arguments point to the lack of
coordination between recurrent and development budgets.

A less often mentioned problem in these circumstances was double or overlapping

budgeting for the same or similar recurrent activities of government. Frequently, line
ministries managed to receive funds from two sources for the same cause by negotiating
separately with the finance and planning ministries. They succeeded in this not only
because development budgets contained substantial non-investment and non-project-
related recurrent spending, in addition to capital spending, but also because the finance
ministries had no proper information, either on the recurrent cost implications of the
investment projects or on the independent recurrent expenditures being financed by
donors—a problem that continues to date in most developing countries.

In several countries, dual budget preparation in terms of contents, classification, and even
calendars, resulted in two different budgets being presented to parliament. Even in some
countries where the timing of budget presentation has been unified, the documentation,
classification, and presentation format of the two budgets has remained separate.
Typically, recurrent budgets were classified on the basis of line ministries and/or those
ministries’ programs and expenditure items, while budgets for capital projects and

15
In some countries, although the minister of finance may have presided over the planning commissions or
planning ministries, the latter were directed from the presidency. For details see I. Linert and F. Sarraf,
“Systemic Weaknesses of Budget Management in Anglophone African Countries”, IMF Working Paper,
2001, IMF Website.

6
developmental recurrent spending were presented as a list of projects, mostly with
identification of their funding sources, but with little or no expenditure itemization.

The main impact of a dual budgeting system on budget execution and reporting occurred
through capital or recurrent projects financed by donors, which usually required
governments to keep separate accounts for domestic counterpart funds and to deposit
such funds along with donor funds in special bank accounts, outside the purview of
central treasuries or accountants-general. These special accounts, mainly in commercial
banks, were governed by flexible spending rules and often bypassed the national budget
systems, undermining the objective of comprehensive government accounting, banking,
and cash management coverage. Moreover, to ensure the availability of funds for
recurrent expenditures of selected capital projects, government authorities created
expedients such as extra-budgetary funds and trust accounts, whose transactions were
subject to separate accounting and banking rules. This contributed to weaker reporting
and delays in closing government accounts, due to the difficulty of reconciling
government accounting records with government banking transactions. Although in most
countries central treasury offices and accountants-general issued circulars requiring line
ministries to report their transactions, in practice these reports were not complete or
timely.

Special project management units were created in line ministries, dealing with the
accounting and banking of the cash components of external loans and grants, as well as
with government counterpart funds. Since these recurrent and capital transactions were
not classified in terms of the object of expenditures, in practice the decision to itemize
expenditures was entrusted to line ministries in the course of budget execution. Several
line ministries did not use the government standard object classification that was used for
traditional recurrent expenditures, thereby multiplying the accounting and reporting
problems.

Late 1990s to date: PRSP, MTEF, sector-wide approach, and budget support era

Since the late 1990s, new financial assistance policies and instruments such as PRSPs,
monitoring of poverty-reducing expenditures, sector-wide and budget support approaches
have been changing the dynamics of donor/client relationships. Moreover, responding to
urgent needs in health, education, and other sectors in poor countries, several donors have
substantially increased their financing of recurrent expenditures. The new approaches
have implications for consultations, conditionality, and data requirements, and for
technical assistance to support an integrated approach to recurrent and development
outlays. Public expenditure reviews (PERs) clearly have moved toward more intra-and
inter-sectoral analyses, combining both recurrent and development expenditures, and
where data are available, reviewing the joint impact of both types of spending on
budgetary outcomes including economic growth, poverty reduction, and asset
maintenance. But while most developing countries have at least begun using medium-
term expenditure frameworks (MTEFs) to replace five-year national plans and public
investment programs, their annual programming and budgeting of recurrent and
development operations still tend to be carried out by separate organizations.

7
Some developing countries have merged their finance ministries with their planning or
development ministries, though others are slow in doing so. The organizational
separation of planning and finance ministries has become institutionalized in culture and
politics in several countries. Political concerns about eliminating a ministerial portfolio,
combined with institutional revivalism between the two ministries at the institutional
level, persist in many developing countries. On the other hand, in considering eliminating
a ministerial portfolio, it should be kept in mind that governments can use their authority
to create as many cabinet posts as they wish—say, for example, creating a ministerial
post for environment or splitting the ministry of trade and industry into two.

Over the years, staff of planning or development ministries have tended to be better paid,
and to benefit from various allowances and better professional treatment, than staff of
finance ministries. Their favored position in part reflects the additional financing
possibilities associated with externally financed projects. This feature might be a cause of
difficulty in the integration of the two institutions.

Line ministries have a strong incentive to prepare and defend separate budgets because
they can use the opportunity of donor-negotiated projects to demand further
complementary financing and to expand their operations without attention to the future
cost implications. These projects may include either capital projects or specific and
independent recurrent activities. For line ministries, dealing with two separate central
ministries for defending their budgets is more advantageous than dealing with one unified
central budget authority, because they can take advantage of the two central ministries’
lack of detailed information.

Today more than ever, the integration of recurrent and development budgets in
developing countries has become a necessity:

• government borrowing is no longer limited to capital expenditures;

• only an integrated analysis of recurrent and development expenditures can identify
those poverty-reducing expenditures that have immediate impact (e.g. social transfers
and targeted subsidies, and some other social expenditures) or an indirect impact
through accelerating economic growth;
• part of external concessional credits, and the bulk of donor grants, are used for
recurrent expenditures, supplementing the government recurrent budget; and
• even after several decades of experience, the recurrent costs of capital projects
continue be ignored, in part for lack of coordination between two separate budgets.

3. Budget integration: four necessary elements

A full and meaningful integration of recurrent and development budgets in developing
countries requires:

• organizational and staffing integration,

• integrated budget preparation,

8
• unified budget documentation and presentation, and
• unified accounting and reporting systems.

This chapter discusses each of these four elements in turn, drawing on lessons of
experience. Budget integration is an area where developing countries can learn more
from each other than from advanced countries. Given that advanced countries integrated
their recurrent and capital budgets decades ago, and that the donor factor does not apply
to them, the recommendations below do not draw on advanced countries’ experience.

Organizational and staffing integration into a single ministry

Medium-term sectoral policy analyses, and their related projections of both recurrent and
development expenditures, are tightly interrelated with the annual budgeting process that
translates them into operational programs and related appropriations. Accordingly, any
organizational separation of these two tasks, either in developing a medium-term
framework or in annual budgeting, can damage their comprehensiveness and quality. The
necessary relationship between sectoral analysis and the annual budget appropriations, on
the one hand, and coordination between recurrent and capital expenditures on the other,
can be maintained if staff dealing with the MTEF and annual budgeting in the central
budget authority and line ministries treat these functions as a unified task.

Ensuring the complementarities between recurrent activities and investment projects,

combined with measuring their outcomes, requires a comprehensive and holistic
approach to initiating, analyzing, and financing government programs. The approach
needs to encompass both the recurrent and capital expenditures of a program or, in the
absence of a program structure in the budget, the expenditures within interrelated
administrative divisions of a line ministry. It requires full organizational integration, not
only shared analysis techniques with separate staffing. In fact, this integration should
permit some savings of scarce expertise in policy analysis and budgeting.

Line ministries should adopt the same pattern, in order to avoid any dual budgeting
before their budget proposals reach the central budget authority. Most external donors
initially approach the line ministries for possible funding of their recurrent and/or capital
expenditures before the proposals, along with the ministry’s traditional recurrent budget,
arrive in the central budget authority for further analysis and finalization. For this reason,
the integration of recurrent and development budgets, as well their integration with
medium-term expenditure planning, are as critical in the line ministries as in the central
budget authority.

In both advanced and developing countries, decisions on large infrastructure projects may
be made at the center of government (presidency, cabinet of ministers, and even by the
legislature) rather than by a finance and planning ministry. But this practice does not
justify the retention of a separate ministry or a planning commission. It does not alter the
need to combine the preparatory work on such projects (feasibility study, appraisal,
annual budgeting, etc.) with government mainstream mid-term expenditure planning and
the use of an annual budgeting system, to ensure the provision of operational and

9
maintenance costs after completion of the capital works. Nor does it remove the need to
establish a large project’s relations with the government’s budget preparation cycle and
other ongoing recurrent activities and capital projects of the government.

It seems that only through an organizational merger can both large capital projects (even
those that are determined at the center of government) and the rest of the budget
(normally finalized by planning ministries and donors in conjunction with line ministries)
be coordinated with the government recurrent budget in an adequately informed
environment. In a single integrated ministry, staff will have better information on the
recurrent operations and domestic financing components of externally financed projects
and will be able to harmonize that information with the rest of the budget (i.e.
domestically financed recurrent spending as well as spending for capital projects). The
integration is also critical to avoid double budgeting, given the rapidly growing share of
external aid that takes the form of recurrent developmental expenditures that supplement
traditional recurrent expenditures but remain as part of the development budget. Many of
these latter recurrent operations, like capital projects, may well have lasting cost
implications after donor funding is discontinued.

Some developing countries have overcome major political and bureaucratic disincentives
and integrated their finance and planning or development ministries. 16 Several English-
speaking developing countries have merged their finance and planning or development
ministries 17, and, interestingly, most French-speaking African countries have completed
this stage. 18 Several middle-income countries, including Mexico, Paraguay, Sri Lanka,
and Thailand, as well as some Eastern European countries such as the Czech Republic,
have closed their planning or development ministries and absorbed them within their
finance ministries.

Nevertheless, several other developing countries, mainly in the Middle East and Africa,
and the former Soviet union, still retain two separate ministries: one for the preparation of

16
For information on country experiences and practices, aside from examining the countries’ budget
documents in the IMF’s Fiscal Library, the World Bank and IMF websites, and some relevant countries’
websites, consultations were undertaken with some team leaders and members of World Bank missions
who had been involved in technical assistance work in some of the countries named in this paper.
17
Bangladesh (Ministry of Finance and Planning), Botswana (Ministry of Finance and Development
Planning), Ethiopia (Ministry of Finance and Economic Development), Jamaica (Ministry of Finance and
Planning), Lesotho (Ministry of Finance and Development Planning), The Gambia (Department of State
for Finance and Economic Affairs), Uganda (Ministry of Finance, Planning and Economic Development).
Tanzania (Ministry of Finance). Tanzania has merged planning and budgeting function of the development
expenditures in the Budget Department of the Ministry of Finance.
18
Cameroon (Ministère de l'Economie et des finances), Cape Verde (Ministère des finances et de la
planification), Central African Republic (Ministère de l'économie, des finances, du budget et de la
coopération internationale), Comoros ( Ministère des finances, du budget et de l'économie), Guinée
(Ministère de l'economie et des finances), Guinée Bissau (Ministère de l'Economie et des finances de
Guinée Bissau), Madagascar (Ministère de l'Economie, des finances et du Budget), Niger (Ministère des
finances du plan et de l'Economie), Rwanda (Ministère des finances et de la planification économique),
and Togo (Ministère de l'Economie, des finances et de la privatisation).

10
the recurrent budget (normally the finance ministry), and another ministry, commission,
or independent organization for the preparation of the development budget, mostly along
with the mid-term expenditure planning function.

The way forward

After the principle of organizational integration is accepted in a country, several options

may be considered for its implementation:

• In a small country, it is most suitable to integrate the macroeconomic and macro-

fiscal analysis tasks, along with aid coordination, sectoral analysis, MTEF, project
appraisal, annual budgeting, and treasury functions, in a ministry of planning and
finance (or a ministry with an equivalent title).

• In a medium-size country, the macroeconomic analysis function may be entrusted

to a small organization outside the ministry of planning and finance, but the
remaining tasks, including macro-fiscal analysis and preparation of the resource
envelope, need to be consolidated in the ministry of planning and finance.

• In large countries, macro-fiscal analysis, aid coordination, sectoral analysis,

MTEF, project appraisal, and annual budgeting functions can form the
responsibilities of a ministry of planning and budgeting, while a separate ministry
of finance may take charge of treasury functions, along with the traditional
oversight of revenue authorities. 19

Integration of budget preparation

Having accomplished the major and fundamental step of merging the finance and
planning or development ministries, the next step would be to bring together the budget
preparation staff in a single budget department supervised by a single manager. The
merger of finance and planning or development ministries, though a necessity, would not
eliminate dual budging unless the recurrent and development budgeting functions are
fully integrated within the new ministry. Experience (e.g. from The Gambia, Lesotho,
and Madagascar) shows that unless this happens, practice will revert to coordination
between two budgets, which has never worked, rather than achieving budget integration.

Even if a single budget department is created for the two functions, real integration will
not take place until the same staff are made responsible for both medium-term
expenditure planning and annual budgeting in the sectors covering both recurrent and
development expenditures. Lesotho, for example, merged its finance and development
and planning ministries so that only one minister and one permanent secretary are in
charge of both budgets, but the staff dealing with the development budget have remained

19
The term “treasury functions” is used here for the usual government accounting, payment, and cash and
debt management tasks, rather than, as in a few countries ( Australia, Canada, and New Zealand), for
government fiscal policy and budgeting tasks.

11
as separate from the recurrent budgeting staff as they were under two ministries; as a
result, real integration has not yet happened.

Several factors, mostly pertaining to institutions and incentives, pose challenges to the
unified budget preparation process. Politicians normally prefer to focus separately on
capital spending because this creates physical assets that they can show to their
constituents (even if they know the assets may not provide services). In their eyes, the
recurrent budget simply disappears into a bureaucratic machine, which they do not trust.
Moreover, the old paradigm of the superiority of public investment spending over current
spending still retains a grip on many players’ minds. Even if donors are increasing their
support to necessary and neglected recurrent expenditures, governments, to the extent
they can, still give priority to new capital spending over maintaining existing assets and
running current services.

Development expenditures are not subject to a ceiling, and at any time, donor-supported
activities may be initiated with or without government counterpart funds. This suits the
line ministries’ interests, and helps maintain the status quo. 20

Recurrent expenditures are mostly locally financed, although this practice is changing
fast due to their increased financing by donors, and to increased general budget support,
which combines recurrent and development budgets. Domestically financed recurrent
operations are unreliable, being vulnerable both to across-the-board cuts in budget
preparation and to shortfalls in cash releases. Development projects (either capital or
recurrent), on the other hand, are largely donor-financed, and their funding is more
assured. For the same reason, donors prefer to fund operational costs within capital
projects, or independent recurrent expenditures in critical sectors, because they, too, have
little faith in the provision of funds through domestic budgets.

As mentioned above, in some countries, even after the merger of the finance and planning
or development ministries, organizational disincentives such as dealing with the staff who
were engaged separately in recurrent and development budgets, conspire to keep the
status quo of separate budget preparation.

These mostly domestic institutional factors may have militated against the integration of
the recurrent and development budgets in the budget preparation phase. Donors do not
normally suggest that the projects that they are prepared to support should remain outside
a country’s budget preparation process, though this may not apply to the budget
execution and accounting phase described later. It is obvious that when the budget is
finalized, it should incorporate all projects agreed with donors. If new donor-assisted
projects are negotiated and started in the course of the year, an aid-receiving country
should incorporate them in a revised or supplementary budget document (like

20
Government counterpart funds are the only portion of the development budget in which the finance
ministries are engaged during budget negotiations with line ministries. As a result, those portions that are
financed through external grants and loans remain almost unanalyzed by the recurrent budget departments.

12
government’s own-financed recurrent budget), submitted to the legislature, so that budget
integrity will be preserved.

It appears that only a few developing countries have advanced in fully integrating their
budget preparation processes. Tanzania and Uganda have made some progress by
integrating the preparation of recurrent and development budgets within a single
organizational unit. Aside from OECD countries, several middle-income countries
including Brazil and some East European countries such as the Czech Republic,
Macedonia, and Slovenia have already completed this stage of integration but, as
mentioned earlier, because of different conditions, in this paper examples have been
chosen from low-income countries.

The way forward

To accomplish the integrated preparation of budgets in a country that has already merged
its finance and planning ministries, it seems that the only viable alternative would be to
create sectoral divisions within a unified budget department. Each division manager and
his/her staff collectively would be responsible for sectoral policies, program analysis, and
annual budgeting. Such informed and focused sectoral divisions would then deal with
line ministries’ recurrent and development operations, both in the medium-term
expenditure framework and annual budget.

In this proposed set-up, external donors’ sector-specific projects should be coordinated

by the budget departments rather than by a separate aid management unit, and the line
ministries would need to work through the budget department for their budget proposals,
whether or not they wish to obtain counterpart funds for their externally financed
operations. At present, the separate departments for the recurrent and development
budget are not fully aware of all the line ministries’ externally financed activities.
Moreover, most aid coordination units as presently constituted deal only with the
administrative formalities of external aid, and once a donor-supported project is
approved, the line ministries neglect to supply them (or budget departments and even
central treasuries) with timely financial data. 21

The number of sectoral divisions appropriate for a budget department depends on a

country’s size and its organizational and human resources. Also, depending on the size of
a country and number of divisions, the responsibility for macro-fiscal analysis and for

21
There is no intention to address aid coordination issues and problems in this paper, but it needs to be said
that aid coordination units have rarely succeeded. For one thing, they do not have full coverage and detailed
information on external inflows. In some countries, donors contact different central agencies, aside from
line ministries. For example, in The Gambia, UNDP works with the President’s Office, while some
bilateral donors contact the Department of State for Foreign Affairs, and other donors normally work with
the Department of State for Finance and Economic Affairs. In Kenya, the aid coordination unit has
minimal or no interaction with the staff who prepare the separate recurrent and development budgets in two
ministries. A government’s need to carry out the administrative work associated with external aid should
not detract from the budget department’s substantive involvement in the details of externally supported
operations.

13
resource envelope determination can be placed under the same budget director or in
another department within the merged ministry.

Integration of budget documentation and presentation

The integration of documentation and presentation of the recurrent and development

budgets—as distinct from the integration of budget preparation—has gained increasing
acceptance in some developing countries. It seems that this element of integration can
proceed somewhat independently from the integration of the budget preparation system
itself.

Some countries have come to accept that better budget documentation is helpful in many
respects, including to show that they are initiating some reforms as well as to facilitate
their working relations with internal and external stakeholders. Reformed budget
classifications, along with computerized budget documentation and printing, have played
an important role in improving budget documentation.

Unification of budget classifications and documentation, though necessary for a full

integration of recurrent and development budgets, does not mean that the two budgets
themselves are prepared and implemented in an integrated manner. But for those
countries that have achieved it, integrated documentation provides a platform from which
to begin a full integration of the two budgets.

In recent years some countries have assembled their separately prepared recurrent and
development budgets in unified documents for presentation to their legislatures. Jamaica,
Namibia, Papua New Guinea (PNG), Sierra Leone, The Gambia, and a few others
provide good examples. Countries including Jamaica, PNG, and Sri Lanka—and, since
the 2004 budget, The Gambia—also use a reformed, unified object/economic
classification for their recurrent budget and development projects under line ministries’
administrative units and their programs. 22

In other countries (e.g. Guyana and Lesotho), the budget document is unified but has two
separate sections: the recurrent budget classified on a program and expenditure object
basis, and the development budget classified on a program basis. In some countries, the
documentation is even less useful. Nigeria, for example, produces a unified budget
document, but its development budget is only identified by one line item per project and
in a separate section from the recurrent budget. Several other countries still prepare and
submit two budget documents to parliament, in most cases with no proper expenditure
object or program classification for development projects. 23 Both Tanzania and Uganda
22
For more information on different types of budget classifications, see: Guidelines for Public Expenditure
Management, International Monetary Fund, 1999, IMF Website, PP.10-11.
23
The necessity of identification of expenditure object classification for development projects has two main
reasons: (i) it is mostly forgotten that a capital project has its own object classification such as feasibility
study, purchase of land, site development, purchase of equipment, construction, etc; and (ii) today, in
several developing countries, most development projects are no longer of capital work nature, but simply
recurrent operations financed by donors or earmarked revenues, which are normally retained and spent by

14
have integrated the preparation of recurrent and development budgets in many respects,
but still document the two, and present them to parliament, separately.

The way forward

Clearly, the direction should be to unify both the classification and documentation of
recurrent and development budgets. This may be facilitated by simple and country-
capacity-based computerized budget documentation, as well as by reforming budget
classifications and through country-capacity-based integrated financial management
information system (IFMIS) projects. 24

Unified budget execution, accounting, and reporting systems

The integration of these systems is a critical part of integrating the budget management
process in developing countries, and perhaps the most difficult. Almost all developing
countries still use different accounting and reporting systems for their recurrent and
development budgets, partly in reflection of the separate budget preparation processes,
separate budget and accounts classifications, and separate banking arrangements they use
for the two budgets. In addition to domestic incentives, donor practices have played an
important role in this continued dualism. The latter are now geared to change, however,
and the use of country systems is gaining ground among donors and country authorities. 25
Achieving this fourth element of integration depends on accomplishing the integration of
budget documentation and classifications, on which certain countries have already made
progress, as just noted.

Overall, no significant attempts have been made to integrate budget execution,

accounting, and reporting systems in developing countries.

First, as mentioned earlier, only in a few countries does the documentation and
appropriation structure for development projects go beyond a single line item for each
project, regardless of the size of the project and the nature of its operations (recurrent or
capital). 26 This state of affairs provides full flexibility to the executive branch (and

line ministries or they are allocated to extra-budgetary funds, such as road or hospital funds that often do
not follow government standard budget classifications.
24
Note, however, that an IFMIS project by itself may not facilitate recurrent/capital budget integration, nor
improve the accounting and reporting system, if the country’s budget and accounts classification is not
reformed and especially if the project is over-sophisticated and loaded with marginal features not adapted
to a country’s capacity to maintain it.
25
See Joint Declaration of Aid Effectiveness, Ownership, Harmonization, Results, and Mutual
Accountability, Paris, March 2005.
26
Other project-related information might be found in the budget documentation, such as project location,
source of financing (domestic or external and, if the second, loans and/or grants), and even information on
detailed project activities. However, none of these data (though useful) is sufficient for an accounting and
reporting system that can be easily combined with data on government recurrent expenditures, which

15
donors) to decide the expenditure object classification of a project in the course of the
budget execution—flexibility that does not apply to items in the recurrent budget.
Although this initially seems to be a budget documentation issue, in practice it leads to
different accounting and reporting process as well.

Second, only in a few countries (perhaps those that have already unified their
classifications in their budget documents) does the object classification of development
projects (determined in the course of the year) match the classification used in the
recurrent budget.

Third, as noted earlier, separate banking arrangements that exist for development projects
outside the treasury’s mainstream arrangements almost always cause serious problems in
reconciling a government’s accounting records with its banking transactions, as part of
the accounting and reporting processes. Some recurrent expenditures (financed either by
donors or government counterpart funds in the development budgets), along with capital
expenditures, are subject to a more flexible pattern of spending and, at times, different
accounting rules. Moreover, operations financed by the funds spent outside the country
by donors are not reported in a timely manner or using the government’s standard
accounting and reporting classifications.

The separate accounting and reporting practices of development budgets normally reflect
strong incentives of all parties involved in the processes. Country systems that are weak
from the donors’ viewpoint, combined with direct and indirect benefits (in the form of
allowances, facilities, travel per diem, etc.) paid to staff involved in preparing and
implementing development projects, have encouraged the use of more flexible
accounting and payment arrangements that remain outside government mainstream
accounting and reporting systems. Up to now, both donors and line ministries have had
strong incentives to defend these arrangements, though this stance may be changing, and
donors are now considering using countries’ own accounting and reporting systems,
although these systems need to be reformed.

In many countries, the rules for the development budget are more flexible than those for
the recurrent, including carrying forward unused appropriations to the next year. Once a
project is established in the development budget (regardless of its capital or recurrent
nature), it is very difficult to remove it, however slow its implementation might be.
Moreover, in many countries, virement rules are easier for the development budget, even
if an object classification is approved or presented in the budget, though such virements
may require donor approval, which is easier than any virements in the domestically-
financed recurrent budget.

Unfortunately, solutions such as the introduction of a treasury single account for banking
of all government transactions have not moved forward, due to the disincentives
mentioned on the part of line ministries and donor agencies. Moreover, banking

contain a detailed expenditure object classification that provides the basis for a government’s financial
reporting system and its chart of accounts.

16
arrangements are only part of the requirement for a fully integrated budget execution,
accounting, and reporting system.

The way forward

Donor factors play a more significant role here than in the other three aspects of budget
integration. On the positive side, donors’ acceptance of the need to help strengthen and
use country systems for their interventions mainly centers on the accounting and
reporting processes, which may further help this element of integration of recurrent and
development budgets.

Proposed actions in this context in the future should include:

• Unifying the budget preparation process for recurrent and development budgets,
including all donor-funded projects, apart from its own merit and necessity, as a
sound framework for accounting and reporting processes.

• Reforming and unifying recurrent and development budget classifications in the

functional, administrative, and economic classification contexts, and where
possible within a program structure. Action is also needed to unify budget
documentation and presentation, as well as to identify the expenditure object
classification of development projects in the budget documents, as the basis for
their accounting and reporting along with recurrent expenditures.

• Requiring donors to use the client country’s expenditure object classification

when reporting the amounts they have spent outside the country on a project.

• Until the budget execution, accounting, and reporting process are fully integrated,
line ministries should reconcile their accounting books with the banking
transactions of donor-funded projects for the funds that the projects managers
receive and spend outside central treasuries, and report those transactions
promptly to the central treasury offices and accountants-general.

• Finally, the opportunity to implement the recommendations of the March 2005

Joint Paris Declaration of Aid Effectiveness by donors and client countries should
be used to accelerate this phase of the integration of the recurrent and
development budgets.

17